The proposed research is directed toward the comprehensive assessment of firesetting children as well as their parents and family environments and will address four interrelated objectives. One objective is to identify empirically the descriptive and clinical profiles of 220 children who have recently set fires using standardized assessment measures and to compare them with those of 220 matched nonfiresetting children. The samples will be composed of children between the ages of six and 12 years from both clinic (patient) and community (nonpatient) samples. A thorough evaluation of those dimensions proposed in the literature provides the conceptual background for the selection of individual measures which assess child psychopathology, demographic status, parent and family characteristics, and firesetting details in order to document what such children and their families are like. The unique aspects of psychiatric samples will also be highlighted in comparisons with nonpatient samples, as the latter have been ignored in the literature. A second objective is to provide a comprehensive functional analysis of the antecedents, characteristics, and consequences of firesetting. This information will shed light on the motives of firesetters, situational determinants, parental discipline strategies, and general consequences in terms of medical, financial, social, legal, and family effects. A third objective is to identify changes in child, parent, and family characteristics over a three-year follow-up period and to relate these changes to the clinical course of firesetting. This analysis will also examine differences between single-incident firesetters and repeaters so that a prospective clinical picture of recidivists can be developed and later employed in preventive interventions. Specific predictors of the initiation and cessation of firesetting will also be identified. The fourth objective is to develop a risk factor model of firesetting as an outgrowth of the follow-up study. This model draws upon a proposed conceptualization of domains identified as significant in the literature. It is anticipated that each risk factor in the model will be clarified once subjected to a rigorous experimental evaluation. The findings will have preventive, therapeutic, and heuristic implications for the control of childhood firesetting.